High Voltage Direct Current Transmission: 11.0 Historical Background

11
High Voltage
Direct Current Transmission

11.0 Historical Background

Power Transmission was initially carried out in the early 1880s using Direct Current (d.c.). With the
availability of transformers (for stepping up the voltage for transmission over long distances and for stepping
down the voltage for safe use), the development of robust induction motor (to serve the users of rotary power),
the availability of the superior synchronous generator, and the facilities of converting a.c. to d.c. when
required, a.c. gradually replaced d.c. However in 1928, arising out of the introduction of grid control to the
mercury vapour rectifier around 1903, electronic devices began to show real prospects for high voltage direct
current (HVDC) transmission, because of the ability of these devices for rectification and inversion. The most
significant contribution to HVDC came when the Gotland Scheme in Sweden was commissioned in 1954 to be
the World's first commercial HVDC transmission system. This was capable of transmitting 20 MW of power
at a voltage of -100 kV and consisted of a single 96 km cable with sea return.

With the fast development of converters (rectifiers and inverters) at higher voltages and larger currents, d.c.
transmission has become a major factor in the planning of the power transmission. In the beginning all HVDC
schemes used mercury arc valves, invariably single phase in construction, in contrast to the low voltage
polyphase units used for industrial application. About 1960 control electrodes were added to silicon diodes,
giving silicon-controlled-rectifiers (SCRs or Thyristors).

In 1961 the cross channel link between England and France was put into operation. The a.c. systems were
connected by two single conductor submarine cables (64km) at 100kV with two bridges each rated at 80
MW. The mid-point of the converters was grounded at one terminal only so as not to permit ground currents to
flow. Sea return was not used because of its effect on the navigation of ships using compasses. The link is an
asynchronous link between the two systems with the same nominal frequency (60Hz).

The Sakuma Frequency Changer which was put into operation in 1965, interconnects the 50Hz and the 60Hz
systems of Japan. It is the first d.c. link of zero length, and is confined to a single station. It is capable of
transmitting 300 MW in either direction at a voltage of 250 kV.

In 1968 the Vancouver Island scheme was operated at +250 kV to supply 300 MW and is the first d.c. link
operating in parallel with an a.c. link.

In 1970 a solid state addition (Thyristors) was made to the Gotland scheme with a rating of 30MW at -
150kV.

Also in 1970 the Kingsnorth scheme in England was operated on an experimental basis. In this scheme
transmission of power by underground d.c. cable at 200 kV, 640 MW is used to reinforce the a.c. system
without increasing the interrupting duty of a.c. circuit breakers.

The first converter station using exclusively Thyristors was the Eel River scheme in Canada. Commissioned in
1972, it supplies 320 MW at 80 kV d.c. The link is of zero length and connects two a.c. systems of the same
nominal frequency (60Hz).
High Voltage Direct Current Transmission
185

The largest thyristors used in converter valves have blocking voltages of the order of kilovolts and currents of
the order 100s of amperes. In order to obtain higher voltages the thyristor valve uses a single series string of
these thyristors. With higher current ratings required, the valve utilizes thyristors directly connected in parallel
on a common heat sink.

The largest operational converter stations have ratings of the order of gigawatts and operate at voltages of 100s
of kilovolts, and maybe up to 1000 km in length.

The thyristors are mostly air cooled but may be oil cooled, water cooled or Freon cooled. With air cooled and
oil cooled thyristors the same medium is used as insulant. With the Freon cooled thyristors, SF6 may be used
for insulation, leading to the design of a compact thyristor valve.

Unlike an a.c. transmission line which requires a transformer at each end, a d.c. transmission line requires a
convertor at each end. At the sending end rectification is carried out, where as at the receiving end inversion is
carried out.

11.1 Comparison of a.c and d.c transmission

11.1.1 Advantages of d.c.

(a) More power can be transmitted per conductor per circuit.

The capabilities of power transmission of an a.c. link and a d.c. link are different.
For the same insulation, the direct voltage V
d
is equal to the peak value (2 x rms value) of the alternating
voltage V
d
.
V
d
= 2 V
a

For the same conductor size, the same current can transmitted with both d.c. and a.c. if skin effect is not
considered.
I
d
= I
a

Thus the corresponding power transmission using 2 conductors with d.c. and a.c. are as follows.

d c power per conductor P
d
= V
d
I
d

a c power per conductor P
a
= V
a
I
a
cos !

The greater power transmission with d.c. over a.c. is given by the ratio of powers.

0
P
d
= _2__ = /1.414 at p.I unity
P
a
cos ! /1.768 at p.I 0.8

In practice, a.c. transmission is carried out using either single circuit or double circuit 3 phase transmission
using 3 or 6 conductors. In such a case the above ratio for power must be multiplied by 2/3 or by 4/3.

In general, we are interested in transmitting a given quantity of power at a given insulation level, at a given
efficiency of transmission. Thus for the same power transmitted P, same losses P
L
and same peak voltage V,
we can determine the reduction of conductor cross-section A
d
over A
a
.
High Voltage Engineering - J R Lucas, 2001 186

Let R
d
and R
a
be the corresponding values of conductor resistance for d.c. and a.c. respectively, neglecting skin
resistance.

For d.c current = P
V
m

power loss P
L
= (P/V
m
)
2
R
d
= (P/V
m
)
2
.(l/A
d)

For a.c current = P = 2 P .
(Vm/2) cos ! Vm cos !

power loss PL = [2P/(Vm cos!)|
2
Ra

= 2 (P/Vm)
2
.(l/A
a cos
2
!)

Equating power loss for d.c. and a.c.

(P/Vm)
2
.(l/A
d) = 2 (P/Vm)
2
.(l/A
a cos
2
!)

This gives the result for the ratio of areas as
0
Ad = cos
2
! = / 0.5 at p.I. unity
A
a
2 / 0.32 at p.I. 0.8

The result has been calculated at unity power factor and at 0.8 lag to illustrate the effect of power factor on the
ratio. It is seen that only one-half the amount of copper is required for the same power transmission at unity
power factor, and less than one-third is required at the power factor of 0.8 lag.

(b) Use of Ground Return Possible

In the case of hvdc transmission, ground return (especially submarine crossing) may be used, as in the case of a
monopolar d.c. link. Also the single circuit bipolar d.c. link is more reliable, than the corresponding a.c. link,
as in the event of a fault on one conductor, the other conductor can continue to operate at reduced power with
ground return. For the same length of transmission, the impedance of the ground path is much less for d.c. than
for the corresponding a.c. because d.c. spreads over a much larger width and depth. In fact, in the case of d.c.
the ground path resistance is almost entirely dependant on the earth electrode resistance at the two ends of the
line, rather than on the line length. However it must be borne in mind that ground return has the following
disadvantages. The ground currents cause electrolytic corrosion of buried metals, interfere with the operation
of signalling and ships' compasses, and can cause dangerous step and touch potentials.

(c) Smaller Tower Size

The d.c. insulation level for the same power transmission is likely to be lower than the corresponding a.c. level.
Also the d.c. line will only need two conductors whereas three conductors (if not six to obtain the same
reliability) are required for a.c. Thus both electrical and mechanical considerations dictate a smaller tower.

(d) Higher Capacity available for cables

In contrast to the overhead line, in the cable breakdown occurs by puncture and not by external flashover.
Mainly due to the absence of ionic motion, the working stress of the d.c. cable insulation may be 3 to 4 times
higher than under a.c.

187

Also, the absence of continuous charging current in a d.c. cable permits higher active power transfer, especially
over long lengths.

(Charging current of the order of 6 A/km for 132 kV). Critical length at 132 kV 80 km for a.c cable. Beyond
the critical length no power can be transmitted without series compensation in a.c. lines. Thus derating which
is required in a.c. cables, thus does not limit the length of transmission in d.c.

A comparison made between d.c. and a.c. for the transmission of about 1550 MVA is as follows. Six number
a.c. 275 kV cables, in two groups of 3 cables in horizontal formation, require a total trench width of 5.2 m,
whereas for two number d.c. 500 kV cables with the same capacity require only a trench width of about 0.7 m.

(e) No skin effect

Under a.c. conditions, the current is not uniformly distributed over the cross section of the conductor. The
current density is higher in the outer region (skin effect) and result in under utilisation of the conductor cross-
section. Skin effect under conditions of smooth d.c. is completely absent and hence there is a uniform current
in the conductor, and the conductor metal is better utilised.

(f) Less corona and radio interference

Since corona loss increases with frequency (in fact it is known to be proportional to f+25), for a given
conductor diameter and applied voltage, there is much lower corona loss and hence more importantly less radio
interference with d.c. Due to this bundle conductors become unnecessary and hence give a substantial saving
in line costs. [Tests have also shown that bundle conductors would anyway not offer a significant advantage
for d.c as the lower reactance effect so beneficial for a.c is not applicable for d.c.]

(g) No Stability Problem

The d.c. link is an asynchronous link and hence any a.c. supplied through converters or d.c. generation do not
have to be synchronised with the link. Hence the length of d.c. link is not governed by stability.

In a.c. links the phase angle between sending end and receiving end should not exceed 30
o
at full-load for
transient stability (maximum theoretical steady state limit is 90
o
).

Note: = w lc per km (2 x 50)/(3 x 10
5
) rad/km (2 x 180 x 50)/(3 x 10
5
) 0.06
0
/km

The phase angle change at the natural load of a line is thus 0.6
o
per 10 km.

The maximum permissible length without compensation 30/0.06 = 500 km

With compensation, this length can be doubled to 1000 km.

(h) Asynchronous interconnection possible

With a.c. links, interconnections between power systems must be synchronous. Thus different frequency
systems cannot be interconnected. Such systems can be easily interconnected through hvdc links. For different
frequency interconnections both convertors can be confined to the same station.

In addition, different power authorities may need to maintain different tolerances on their supplies, even though
nominally of the same frequency. This option is not available with a.c. With d.c. there is no such problem.

(i) Lower short circuit fault levels

When an a.c. transmission system is extended, the fault level of the whole system goes up, sometimes
necessitating the expensive replacement of circuit breakers with those of higher fault levels. This problem can
be overcome with hvdc as it does not contribute current to the a.c. short circuit beyond its rated current. In fact
it is possible to operate a d.c. link in "parallel" with an a.c. link to limit the fault level on an expansion. In the
event of a fault on the d.c line, after a momentary transient due to the discharge of the line capacitance, the
current is limited by automatic grid control. Also the d.c. line does not draw excessive current from the a.c.
system.

(j) Tie line power is easily controlled

In the case of an a.c. tie line, the power cannot be easily controlled between the two systems. With d.c. tie
lines, the control is easily accomplished through grid control. In fact even the reversal of the power flow is just
as easy.

11.1.2 Inherent problems associated with hvdc

(a) Expensive convertors

Expensive Convertor Stations are required at each end of a d.c. transmission link, whereas only transformer
stations are required in an a.c. link.

(b) Reactive power requirement

Convertors require much reactive power, both in rectification as well as in inversion. At each convertor the
reactive power consumed may be as much at 50% of the active power rating of the d.c. link. The reactive
power requirement is partly supplied by the filter capacitance, and partly by synchronous or static capacitors
that need to be installed for the purpose.

(c) Generation of harmonics

Convertors generate a lot of harmonics both on the d.c. side and on the a.c. side. Filters are used on the a.c.
side to reduce the amount of harmonics transferred to the a.c. system. On the d.c. system, smoothing reactors
are used. These components add to the cost of the convertor.

(d) Difficulty of circuit breaking

Due to the absence of a natural current zero with d.c., circuit breaking is difficult. This is not a major problem
in single hvdc link systems, as circuit breaking can be accomplished by a very rapid absorbing of the energy
back into the a.c. system. (The blocking action of thyristors is faster than the operation of mechanical circuit
breakers). However the lack of hvdc circuit breakers hampers multi-terminal operation.

(e) Difficulty of voltage transformation

Power is generally used at low voltage, but for reasons of efficiency must be transmitted at high voltage. The
absence of the equivalent of d.c. transformers makes it necessary for voltage transformation to carried out on
the a.c. side of the system and prevents a purely d.c. system being used.
189

(f) Difficulty of high power generation

Due to the problems of commutation with d.c. machines, voltage, speed and size are limited. Thus
comparatively lower power can be generated with d.c.

(g) Absence of overload capacity

Convertors have very little overload capacity unlike transformers.

11.1.3 Economic Comparison

The hvdc system has a lower line cost per unit length as compared to an equally reliable a.c. system due to the
lesser number of conductors and smaller tower size. However, the d.c. system needs two expensive convertor
stations which may cost around two to three times the corresponding a.c. transformer stations. Thus hvdc
transmission is not generally economical for short distances, unless other factors dictate otherwise. Economic
considerations call for a certain minimum transmission distance (break-even distance) before hvdc can be
considered competitive purely on cost.

Estimates for the break even distance of overhead lines are around 500 km with a wide variation about this
value depending on the magnitude of power transfer and the range of costs of lines and equipment. The break-
even distances are reducing with the progress made in the development of converting devices.

Figure 11.1 shows the comparative costs of d.c. links and a.c. links with distance, assuming a cost variation of
5% for the a.c. link and a variation of 10% for the d.c. link.

For cables, the break-even distance is much smaller than for overhead lines and is of the order of 25 km for
submarine cables and 50 km for underground cables.

cost/MW

distance (km)
Figure 11.1 - Break-even distance for d.c. transmission
0 100 200 300 400 500 600 700 800 900
80

70

60

50

40

30

20

10
d.c.
a.c.
11.2 Convertor arrangements and operation

The three-phase six-valve bridge rectifier is almost exclusively used in high voltage direct current applications.
This is shown in figure 11.2.

The 6-valve bridge connection gives double the direct voltage as output compared to the simple 3-phase
rectifier. The convertor transformer may either be wound star-star as shown, or as star-delta (or even as delta-
star or delta-delta). The ripple of each of these connections is the same, but are phase shifted by 30
0
in output
with respect to each other. To obtain a smoother output, two bridges (one star-star and the other star-delta) may
be connected together to give the twelve pulse connection.

For the 6-valve bridge, with zero firing delay, the voltage waveforms across the thyristors are shown in figure
11.3. At any given instant, one thyristor valve on either side is conducting. The conducting period for the
thyristor valve R1 is shown on the diagram.
The corresponding d.c. output voltage waveforms are shown in figure 11.4.

It can be shown that for the 6-valve bridge, the total r.m.s. ripple is of the order of 4.2% of the d.c. value (for
zero delay 0 and zero commutation 0). The ripple oI course increases with the delay angle and has a
value oI about 30 at /2.

star-star convertor transformer Graetz bridge

Figure 11.2 - Hexa-valve Bridge Convertor arrangement
R1 Y3 B5 R1

Y6 B2 R4 Y6

Figure 11.3 - Thyristor voltage waveforms ( 0)

Y6,R1 R1,B2 B2,Y3 Y3,R4 R4,B5 B5,Y6 Y6,R1

Figure 11.4 - d.c. output waveforms ( 0)
B
Y
R
1 3 5
4 6 2
choke
+

191

With the 12 pulse bridge, the r.m.s. ripple is of the order of 1.03% of the d.c. value (for 0 and 0), and
increases to about 15 at /2.

The use of a choke reduces the ripple appearing in the direct current transmitted.

If E is the r.m.s. line-to-line voltage, then if 0, 0, then the direct voltage output is given by equation
(11.1).

11.2.1 Control angle (Delay angle)

The control angle for rectification (also known as the ignition angle) is the angle by which firing is delayed
beyond the natural take over for the next thyristor. The transition could be delayed using grid control. Grid
control is obtained by superposing a positive pulse on a permanent negative bias to make the grid positive.
Once the thyristor fires, the grid loses control.

Assuming no commutation (2 thyristors on same side conducting simultaneously during transfer), the voltage
waveforms across the thyristors as shown in figure 11.5.
In this case, the magnitude of the direct voltage output is given by the equation (11.2).

11.2.2 Commutation angle (overlap angle)

The commutation period between two thyristors on the same side of the bridge is the angle by which one
thyristor commutates to the next. During this period , the voltage at the electrode Iollows mean voltage oI the
2 conducting thyristors on the same side. This is shown in figure 11.6.

3
x 2
3
1
.
2 3
. E =
d
2
3
x 2 x
3
E
x 2 =
V
3
3

-
do
1
]
1
sin
cos

E 1.350 = E .
2 3
=
Vdo
(11.1)
R1 Y3 B5 R1

Y6 B2 R4 Y6

Figure 11.5 - Thyristor voltage waveforms (with delay )

-
3
+ +
3

3
1
.
2 3
. E =
d
2
3
x 2 x
3
E
x 2 =
V
+
+ -
d
3
3
1
]
1
,
_
,
_
sin sin
cos

V
= E
2 3
=
V do d

cos cos
(11.2)

With both the delay angle and commutation being present, the magnitude of the direct voltage may be
determined from equation (11.3) as follows.

An alternate method of derivation of the result is based on comparison of similar areas on the waveform.
Figure 11.7 gives the necessary information.
d.c. output = average value of waveform

R1 Y3 B5 R1

Figure 11.6 - commutation between 2 thyristors
( )
1
1
]
1
d . + d . + ) + (
3
E 2 3
=
d ) ( f
2
3
2
3
E
2 =
V

+
+ + -
3
2
2
1
+ + -
+ -
+
+ -
d
3
3
3
3
3
3

cos cos cos

[ ] ) + ( +
2
V
=
V

do
d
cos cos
(11.3)

Figure 11.7 - Graphical analysis of waveform

d . ) ( V
3 / 2
1
=
V
3
2
+ +
+
d

w
A
1
B
2
A
2
B
1
B
3
A
3
0 + 2/3++ + ++
193

In this integral, in graphical form, area A
1
can be replaced by area B
1
. Similarly, area A
2
can be replaced by
area B
2
and area A
3
by area B
3
. The integral equation then reduces to the form shown below.

where 2 E is the peak value of the line voltage.

Simplification gives the desired result as in equation (11.3).

Commutation is a result of the a.c. system inductance L
c
, which does not allow the current through a thyristor
to extinguish suddenly. Thus the larger the current the larger the commutation angle .

If the extinction angle for rectifier operation w is defined as w = , then V
d can be written as in equation
11.4.

Consider the commutation between thyristors 1 & 3 in the bridge circuit shown in figure 11.8. Let ic be the
commutation current. The commutation current is produced by the voltage ec which is the line voltage between
phases 1 and 2.

This can also be written in the following form, with ic = 0 at the start of commutation and ic = Id at the end of
commutation.
[ ] ) - ( - ) + (
2
E 2 3
=
d
2
E 2 3
=
V

-
+
d

cos cos
sin

[ ]
[ ] ) + ( +
2
V
=
) + ( +
2
E 2 3
=
V
0
d

cos cos
cos cos

Figure 11.8 - Circuit for analysis of commutation

[ ] w + =
V
2
1
d
cos cos
(11.4)

t E 2 =
)
3
- t ( - )
3
+ t (
3
2 E
=
E
-
E
=
e 1 2 c
sin
cos cos
1
]
1

1 3 5
4 6 2
+
Lc
E1
Lc
E2
ec
ic
Ld Id

This equation may be integrated as follows to give equation 11.5.

Equation 11.4 can also be written of the form of equation (11.6).

From equations (11.5) and (11.6) w may be eliminated to give the equation (11.7).

It can also be shown that equation (11.7) can be rewritten in terms of the extinction angle w instead of the
ignition angle as in equation (11.8).
Equation (11.7) suggests that the convertor may be represented by a no load d.c. output voltage V
o
cos
and a series fictitious resistance
c
/ as shown in figure 11.9. The fictitious resistor does not however
consume any active power.

t d
i
d

L
2 =
e
c
c c
i
d
L e c
I
0
c c
+ = w
2
1
d
= ) t ( d

I L
= ) w - (
2
E
i.e.
I L
= ) t ( d t E 2
d c
d c
w

2
1

cos cos
sin

I
E
L
2
= w -
d
c
cos cos
(11.5)

V
V
2
= w +
o
d
cos cos
(11.6)

V
L
3
+
V
V
= ,
E 2 3
=
V
I
E 2
L
+
V
V
=
0
c
0
d
0
d
c
0
d
cos
cos

I
L
3
-
V
=
V

d
c
0 d
cos
(11.7)
I
L
3
+ w
V
=
V

d
c
0 d
cos
(11.8)

R = 3Lc/ Id

] ]Q
] ''
] ]
] ]
Vo cos o + ]
] Vd
] ]
] ]
] ]
'Q
Figure 11.9 - Equivalent circuit for Rectifier
195
11.2.3 Current Waveforms

If Commutation is not considered, the current waveforms through each thyristor (assuming a very high value of
inductance L
d
in the d.c. circuit to give complete smoothing) is a rectangular pulse lasting exactly one-third of a
cycle. This is shown in figure 11.10 for the cases without delay and with delay.
When commutation is considered, the rise and fall of the current waveforms would be modified as they would
no longer be instantaneous, as shown in figure 11.11.
Since each phase has 2 thyristors on the opposite half cycles, the a.c. current waveform on the secondary side
of the transformer has a non-sinusoidal waveform as shown in figure 11.12.

If commutation angle is not considered, we can easily calculate the r.m.s. value of the a.c. current on the
transformer secondary I
s
as in equation (11.9).

V
1
, , V
1

I
1

I
d

0 with
Figure 11.10 - Thyristor current waveforms ()
V
1
V
1

2/3
I
1
I
1

2/3

Figure 11.11 - Thyristor current waveforms
V
1
V
1

I
1
I
1

/6 5/6 , , 2/3 |

0, 0 with and

Figure 11.12 - Current waveforms on transformer secondary

Usually harmonic filters are provided on the a.c. system, so that only the fundamental component need to be
supplied/absorbed from the a.c. system. From Fourier analysis, it can be shown that the fundamental
component is given as follows, resulting in equation (11.10).

If filters were not provided, it can be shown, using the Fourier series analysis, that the r.m.s. ripple on the a.c.
system would be 0.242 I
d
(or 31 % of the fundamental)
Note: For normal operation neglecting the commutation angle, in the above calculations of the alternating
current, gives rise to an error only of the order of 1%.

As can be seen from the voltage and current waveforms on the a.c. side, the current lags the voltage due to the
presence of the delay angle and the commutation angle . Thus since only active power is transmitted in a
d.c. link, the convertor must consume the reactive power.

11.2.4 Power factor cos !

Since the convertor consumes reactive power, there will be a power factor associated with the convertor on the
a.c. side. This can be calculated as follows.

Active power supplied to d.c. link = V
d
I
d

Active power supplied from a.c. system = 3 E I cos !

Since the convertor does not consume any active power, there must be power balance.

V
d
I
d
= 3 E I cos !

From this the power factor can be calculated as follows.

This gives the result as in equation (11.11).
cos ! (cos cos w) |cos cos()| (11.11)
In the absence of commutation, this reduces to the simple relationship
cos ! cos
Which means that is the power Iactor angle in the absence oI commutation. The presence oI commutation
reduces the effective power factor by increasing the effective angle.

I 0.8165 = I
3
2
d
2
s
d d =
I
.
3
2
.
1
=
I
(11.9)

) t ( d t
I
.
2
.
2
1
=
2
I
= I
d

-
3
3

cos
max

I
0.78 =
I
6
=
3
2
I
2
= I
d d d
sin
(11.10)

I
V
2 3
3
I
6
w) + (
V
2
1
=
I E 3
I V
=
0
0
d d
cos cos
cos
cos

197

With 0, the active power transmitted is 3 E I cos , and is zero in value when 90
o
.

Thus if a high inductance is connected with the load, the limit of power transfer under rectification is 90
o
.
However, if no inductance were present with the load ( i.e. Ld=0) then the voltage and the current waveforms
would become identical in shape (since the load is purely resistive). Under these conditions, the voltage cannot
go negative at any instant of time, since the current cannot flow in the opposite direction through the thyristors.
The power transmitted would then become zero only at 120
o
. Figure 11.13 shows typical output current
and voltage waveforms.

11.2.5 Current waveforms on a.c. system

In section 11.3.3 it was seen that the current waveform on the secondary side of the convertor transformer was
as shown in figure 11.14
The same current waveform appears on the primary side of the convertor transformer (a.c. system) when the
winding connection is star-star, if no filters are present. If filters are present they would filter out some of the
harmonics and make the a.c system currents more or less sinusoidal.

When the convertor transformer is delta-star connected, the corresponding primary current waveforms would
be as shown in figure 11.15.

Each of the waveforms shown in figures 11.14 and 11.15 contain the same harmonics (6k1) with their
amplitudes inversely proportional to the harmonic number. The polarities of some of the components of the
two waveforms are opposite, so that when a star-star and delta-star convertor bridge are connected in cascade,
only the (12k1) harmonics remain. Thus less filtering is required. Further when commutation is present, the
harmonic amplitudes reduce due to smoothing action.
Vd Vd

Id Id

L
d
= L
d
= 0

Figure 11.13 - Typical d.c. output waveforms
I
d

-2/3 2/3
-/3 0 /3
-I
d

Figure 11.14 - Current waveform on a.c. system
2I
d
/3
I
d
/3
- 0

-2I
d
/3

Figure 11.15 - Primary current waveform

Note: During normal operation of the convertor as a rectifier, the delay angle is small (about 10
0
), while the
commutation angle is comparatively larger (about 20
0
).

11.2.6 Inversion

Because the thyristors conduct only in one direction, the current in a convertor cannot be reversed. Power
reversal can only be obtained by the reversal of the direct voltage (average value) V
d
.

For inversion to be possible, a high value of inductance must be present, and the delay angle ~ 90
0
, since Vd
changes polarity at this angle. The theoretical maximum delay for inversion would occur at 180
0.
Thus it is common practice to define a period of advance from this point rather than a delay from the previous
cross-over as defined for rectification. Thus we define - as the ignition angle for inversion or the angle
oI advance. Similarly the extinction angle is also deIined as - w. The definition of the commutation
angle is unchanged. Thus .

It must be noted, that unlike with rectification which can be operated with 0, inversion cannot be carried
out with 0, since a minimum angle
0
is required for deionisation of the arc and regaining grid control.

Thus we have the practical relationship
0
< /2. Practical values oI
0 lie between 1
0
and 8
0
.

Inversion cannot of course be carried out without a d.c. power source. Further, to obtain the necessary
frequency for the a.c. on inversion, the commutation voltage is obtained from either synchronous machines or
from the a.c. system fed. In isolated systems, L C circuits may also be sometimes used for the purpose. Figure
11.16 shows the thyristor voltage waveforms for inversion.

During inversion, each thyristor conducts during the negative half cycle, so that the direct voltage waveform
and the corresponding current has the form shown in figure 11.17.
R4 Y6 B2

B5
R1 Y3

Figure 11.16 - Thyristor voltage waveforms for inversion

Figure 11.17 - Direct voltage waveform & thyristor current waveform
199
The equations derived earlier for the convertor are valid. However, they are usually written in terms of the
variables and instead oI and w. Thus the equation V
d
= V
0
[cos cos ()| becomes

V
d
= V
0
[cos ( - ) cos ( - )|

or (-) V
d
= V
0
[cos cos | V
0
cos (3L
c/) I
d

Since the direct voltage is always negative during inversion, it is common practice to omit the negative sign
from the expression.

It can also be shown that
(-) Vd = V0 cos - (3L
c/) I
d

As in the case of rectification, it can be shown that for inversion

cos - cos 3L
c/E . Id and I = 6/ I
d

The power factor of the invertor can be shown to be given by the equation

cos ! (cos cos )

It is common practice to operate the invertor at a constant extinction angle (10
0
to 20
0
).

11.3 Control Characteristics

The control characteristics of the convertor are the plots of the variation of the direct voltage against the direct
current. These are described in the following sections.

11.3.1 Natural Voltage Characteristic (NV) and the Constant Ignition Angle (CIA) control

The Natural Voltage Characteristic corresponds to zero delay angle 0. This has the characteristic equation
given by Vd = V0 - (3L
c/) I
d. The Constant Ignition Angle control is a similar characteristic which is
parallel to the NV characteristic with a controllable intercept V0 cos . These are shown in Iigure 11.18.

11.3.2 Constant Extinction Angle (CEA) control

The Invertor is usually operated at constant extinction angle. This has the characteristic equation given by Vd
= V0 cos - (3L
c/) I
d. This is shown in figure 11.19.
Vd

N.V. Characteristic of Rectifier
Vo
Vo cos

C.I.A. Control

Id

Figure 11.18 - N.V. Ch. & C.I.A. Control of Rectifier

11.3.3 Constant Current Control (CC)

In a d.c. link it is common practice to operate the link at constant current rather than at constant voltage. [Of
course, constant current means that current is held nearly constant and not exactly constant].
In constant current control, the power is varied by varying the voltage. There is an allowed range of current
settings within which the current varies.

11.3.4 Full Characteristic of Convertor

The complete characteristic of each convertor has the N.V. characteristic and equipped with C.C. control and
the C.E.A. control. This is shown in figure 11.20 for a single convertor.

Note: The constant current controller adjusts the firing angle so that the current is maintained at a set value,
even for short-circuits on the d.c. line. The C.C. control is present in the invertor too, although the invertor is
not usually operated in that region. The rectifier is normally operated in the C.C. region while the invertor is
operated in the C.E.A. region.

0
Id

Vo cos C.E.A. control

V
d

Figure 11.19 - N.V. Characteristic & C.I.A. Control of Rectifier
V
d

N.V. Characteristic of Rectifier
Vo
Vo cos

Rectification C.I.A. Control

C.C. control

I
ds
Id

Inversion
Vo cos

Vo cos o C.E.A. control

V
d

Figure 11.20 - Complete characteristic of Convertor
201
11.3.5 Compounding of Convertors

Figure 11.21 shows a system of 2 convertors, connected by a hvdc link. Both convertors are provided with
CEA and CC control so that either can work as a rectifier or an invertor. The compounded characteristics are
shown in figure 11.22.
The margin setting I
dm
between the current setting I
ds
for the invertor and for the rectifier is usually kept at
about 10% to 20% of the current setting. The setting of the convertor operating as rectifier is kept higher than
the setting of that as invertor by the margin setting I
dm
.
The usual operating point for power transfer is the intersection of the CC control of the rectifier and the CEA
control of the invertor. (For comparison, the characteristics of convertor B has been drawn inverted). It must
also be ensured by proper tap changing that the N.V. characteristic of the convertor operating in the
rectification mode is higher than the C.E.A. characteristic of the invertor, as V
o
of the two ends are not
necessarily equal.

Figure 11.21 - Compounding of Convertors

Figure 11.22 - Compounded characteristic of Convertors

C.E.A.
Control
Phase
Advance
Circuit
C.C.
Control

Control for

Convertor B
Control for
Convertor A
hvdc link
convertor B convertor A
3-phase ac
I
d
I
d
I
dm
I
ds
E
V
d
A as Rectifier
B as Invertor
A as Invertor
B as Rectifier
N.V. Characteristic of A
N.V. Characteristic of B
C.E.A. control of A
C.E.A. control of B
C.C. control of A
(A as rectifier)
changed
C.C. control of A
for A as invertor
V
o
V
o
cos

V
o
V
o
cos

V
d
I
ds
+I
dm
I
ds
I
ds
I
dm
With convertor A operating as rectifier, and convertor B operating as invertor, the steady state current under all
circumstances will remain within the upper limit (I
ds
+ I
dm
) and the lower limit I
ds
. That is, the system direct
current will not change by more than I
dm
under all operating conditions. By reversing the margin setting I
dm
,
that is making the setting of convertor B to exceed that of A, power flow can be automatically reversed.
Convertor B will then operate as a rectifier and A as an invertor. The reversal of power occurs as a result of the
reversal of polarity of the voltage.

11.3.6 Per Unit Convertor Chart

The convertor operating equations for voltage V
d
and current I
d
are expressed as follows.

It is useful to draw the convertor chart in per unit. For this purpose the natural selection for the base voltage is
the maximum direct voltage output V
do
. There is no such natural current base. Thus it is convenient to select
the constant appearing in equation (11.5) for current as the base quantity. Thus the selected base quantities are
as given in equation (11.15).

2
) w + (

E 2 3
=
Vd
cos cos

) w - (
L
2
E
=
I
c
d
cos cos

Figure 11.23 - Per unit convertor chart
V
d
I
d
1.0
0.5
0
-0.5
-1.0
0.5 1.0
w =15
o
w =30
o
w =45
o
w =60
o
w =75
o
w =90
o
w =105
o
w =120
o
w =135
o
w =150
o
w =165
o
w =180
o
=15
o
w =0
o
=0
o =30
o
=45
o
=60
o
=75
o
=15
o
=30
o
=45
o
=60
o =90
o
=105
o
=120
o
=135
o
=150
o
=180
o
=165
o
203

With these base quantities, the per unit voltage V
d
and per unit current I
d
are given as follows.

V
d
= ( cos cos w ) and I
d
= cos - cos w

which give the simple per unit characteristic equations (11.16) and (11.17).

V
d
+ 0.5 I
d
= cos (11.16)
V
d
- 0.5 I
d
= cos w (11.17)

In the chosen per unit system, constant delay angle control (or the natural voltage characteristic) correspond to
straight lines with slope (-) 0.5 on the per unit voltage-current chart. Similarly constant extinction angle control
correspond to straight lines with slope (+) 0.5.

For inversion, the corresponding equations are as follows.

V
d
+ 0.5 I
d
= cos ( - ) (-) cos
V
d
- 0.5 I
d
= cos ( - ) (-) cos

Since voltages are measured in the opposite direction for inversion, these equations may be rewritten in their
more usual form as in equations (11.18) and (11.19).

(-) V
d
- 0.5 I
d
= cos (11.18)
(-) V
d
+ 0.5 I
d
= cos (11.19)

From the characteristics, the constant commutation angle characteristic may be obtained by subtraction as
follows.
- or w -

The constant characteristics can be shown to be ellipses ( 60
0
). The per unit operating chart of the
convertor is shown in figure 11.23.

11.4 Classification of d.c. links

D.C. links are classified into Monopolar links, Bipolar links, and Homopolar links.

In the case of the monopolar link there is only one conductor and the ground serves as the return path. The link
normally operates at negative polarity as there is less corona loss and radio interference is reduced. Figure
11.23 (a) shows a monopolar link.

The bipolar links have two conductors, one operating at positive polarity and the other operating at negative
polarity. The junction between the two convertors may be grounded at one or both ends. The ground does not
normally carry a current. However, if both ends are grounded, each link could be independently operated when
necessary. This is shown in figure 11.23 (b).

L
2
E
=
I
,
E 2 3
=
V
c
base base

(11.15)
The homopolar links have two or more conductors having the same polarity (usually negative) and always
operate with ground path as return.

11.4.1 Harmonics and Filters

As was mentioned earlier, the harmonics present on the a.c. system are (6k1). Thus the a.c. harmonic filters
are tuned to the 5th, 7th, 11th, and 13th harmonics to reduce the harmonic content in the voltages and currents
in the a.c. network to acceptable levels. Higher harmonics would not penetrate very far into the a.c. system.
The harmonics are mainly present in the a.c. current as the a.c. voltage is heavily dependant on the a.c. system
itself.

The Harmonics present on the d.c. side are mainly on output voltage. These are in multiples of 6 as the
waveform repeats itself 6 times. The d.c. current is smoothed by the smoothing reactors.

(a) monopolar link

(b) bipolar link

(c) homopolar link

Figure 11.24 - Kinds of d.c. links

High Voltage Direct Current Transmission: 11.0 Historical Background

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

High Voltage Direct Current Transmission: 11.0 Historical Background

Uploaded by

Copyright:

Available Formats

11

High Voltage Direct Current Transmission

cos cos cos

You might also like